Abstract :
Detection efficiencies ε of 3H by 4πβ liquid scintillation (LS) spectrometry systematically vary as a function of both total cocktail mass m and cocktail composition (such as the H2O mass fraction f). Quench indicating parameter (QIP) determinations, such as for the Horrocks number H, are also dependent on m and f. These systematic effects were investigated with a matrix of 33 LS cocktails covering a broad array of m and f values: 3.0 g not precedes, equals m not precedes, equals 21.4 g and 0.002 not precedes, equals f ≤ 0.52. The effects of m and f on the LS background counting rates of matched blanks are also treated in detail. The uncertainties associated with these m- and f-dependent effects on ε and on H are particularly given close scrutiny. The systematics clearly demonstrate that efficiency changes cannot be adequately monitored by quench indicating parameters when the quench changes are induced by multiple causal factors (e.g. simultaneously varying cocktail sizes and compositions). The experimental results presented here have significant implications on the critical importance of understanding the systematics of these variable effects for a given LS system (combination of cocktails and spectrometer), and of very closely matching LS cocktail volumes and compositions in order to achieve reproducible and accurate LS counting results.
